R E G U L A R
A R T I C L E
Juror Knowledge and Attitudes
Regarding Mental Illness Verdicts
Lisa M. Sloat, MD, and Richard L. Frierson, MD
We begin with a brief overview of the Not Guilty by Reason of Insanity (NGRI) and Guilty but Mentally Ill (GBMI)
verdicts in the United States and then report on a study of qualified jurors (n ⫽ 96) in which we examined jurors’
understanding and attitudes about mental illness verdicts and the disposition of mentally ill defendants. Results
indicate that although the jury pool was highly educated, only 4.2 percent of jurors could correctly identify both
the definitions and dispositions of defendants found NGRI and GBMI. Jurors with lower educational levels were
less likely to identify the dispositional outcome of a GBMI verdict (p ⬍ .05). Eighty-four percent of respondents
believed that juries should be informed of dispositional outcome before deciding a verdict. Also, 68.4 percent of
jurors erroneously believed that a defendant found GBMI could not receive the death penalty. Among jurors who
correctly identified the definition of GBMI, those with lower educational levels were more punitive in their
attitudes toward disposition of the GBMI defendants, believing they should eventually be sent to prison (p ⬍ .05).
J Am Acad Psychiatry Law 33:208 –13, 2005
The disposition of mentally ill offenders continues to
be a controversial issue as lawmakers try to balance
the public’s demand for retribution and safety with
the offender’s civil rights and need for treatment. A
large majority of states have enacted legislation that
establishes an insanity defense. These statutes are
based on the common law principle that the accused
should not be held criminally responsible if he or she
was “insane” at the time of the offense. This principle
was summarized by Judge David Bazelon: “Our collective conscience does not allow punishment where
it cannot impose blame” (Ref. 1, p 876). The criteria
used to determine eligibility for a verdict of Not
Guilty by Reason of Insanity (NGRI) vary across
jurisdictions. Most states use some version of the
M’Naghten rule, which requires that defendants be
mentally ill to the extent that they do not know the
nature and quality of their criminal acts or that their
criminal acts are wrong.2 Most of the remaining
states use the American Law Institute’s (ALI) standard, which holds that defendants are not criminally
responsible if, as a result of mental disease or defect,
they lacked substantial capacity either to appreciate
Dr. Sloat is Clinical Assistant Professor of Psychiatry, University of
Maryland Medical School, and Staff Psychiatrist, Clifton T. Perkins
Hospital Center, Jessup, MD. Dr. Frierson is Associate Professor of
Psychiatry, William S. Hall Psychiatric Institute, University of South
Carolina, Columbia, SC. Address correspondence to: Richard L. Frierson, MD, University of South Carolina, William S. Hall Psychiatric
Institute, PO Box 119, Columbia, SC 29203. E-mail: rlf51@dmh.
state.sc.us
208
the criminality of their conduct or to conform their
conduct to the requirements of the law.3 In most
jurisdictions, defendants found NGRI are hospitalized for treatment in a psychiatric hospital and released when they are no longer mentally ill or
dangerous.
In addition to these insanity statutes, several states
have enacted legislation that includes a separate
Guilty but Mentally Ill (GBMI) verdict. The first
GBMI statute was enacted in Michigan in 1975 in
reaction to the Michigan Supreme Court’s decision
in People v. McQuillan.4 Before this decision, insanity acquittees were automatically committed to the
Michigan Department of Mental Health for an indeterminate time. This practice was declared unconstitutional because it violates due process and denies
the insanity acquittees equal protection by failing to
provide them with release procedures available to individuals committed under other civil commitment
statutes. As part of the McQuillan decision, the court
ordered that all previously committed insanity acquittees be evaluated and released if they no longer
met criteria for civil commitment. A total of 270
patients were evaluated and 214 were released.
Within a year of their release, two of the acquittees
committed heinous and highly publicized crimes.5
The Michigan GBMI statute6 was enacted a year
later in response to public outcry. The statute stated
that a defendant could be found GBMI if the trier of
The Journal of the American Academy of Psychiatry and the Law
Sloat and Frierson
fact found that the defendant was guilty of the offense, was mentally ill at the time of the offense, and
did not lack the substantial capacity either to appreciate the nature and quality or wrongfulness of his
conduct or to conform his conduct to the requirements of the law. It further stated that the same criminal sanctions should be imposed on a person found
GBMI as on a person found guilty.
At least eight additional states created GBMI statutes in response to the insanity acquittal of John
Hinkley, Jr., in 1982, after his attempted assassination of President Reagan. Currently, 12 states utilize
some form of GBMI verdict: Alaska, Delaware,
Georgia, Illinois, Indiana, Kentucky, Michigan,
New Mexico, Pennsylvania, South Carolina, South
Dakota, and Utah.7
Mental Illness Verdicts in South Carolina
South Carolina defines insanity (NGRI) as an affirmative defense that a defendant, as a result of mental disease or defect, “lacked the capacity to distinguish moral or legal right from moral or legal wrong
or to recognize the particular act charged as morally
or legally wrong.”8
The GBMI statute was enacted in South Carolina
in April of 1984. The statute states that a person is
GBMI9:
. . .if at the time of the commission of the act constituting the
offense, he had the capacity to distinguish right from wrong or
to recognize his act as being wrong. . .but because of a mental
disease or defect he lacked sufficient capacity to conform his
conduct to the requirements of the law.
Persons adjudicated GBMI must be taken to a facility designated by the Department of Corrections for
treatment and retained there until, in the opinion of
the staff at the facility, such inmates can be safely
moved to the general population of the Department
of Corrections to serve the remainder of their
sentences.
Criticisms of the GBMI Verdict
The American Psychiatric Association,10 American Psychological Association,11 and American Bar
Association12 have all opposed the GBMI verdict.
GBMI statutes have been widely criticized for several
reasons.
One criticism is that persons found GBMI are
subject to the same criminal sanctions as offenders
who are found guilty, including incarceration and
death. Some have argued that the statute may as well
be called “guilty but ignorant” or “guilty but remorseful,” since GBMI does not indicate mitigated
or diminished capacity.13 In fact, studies have shown
that defendants found GBMI receive longer sentences. For example, Callahan et al.14 compared the
sentences of those offenders who pled insanity and
were subsequently found guilty with those who were
found GBMI. Their results indicated that defendants found GBMI often received longer sentences
than those found guilty with no mental disorder.
Specifically, they found that those receiving a GBMI
verdict were sent to prison 80 percent of the time
(rather than receiving probation) as opposed to 62.8
percent of those found guilty without a mental illness. Also, those found GBMI received a life sentence
14 percent of the time as opposed to 5.5 percent of
the time for those found guilty without mental disorder. This difference was even more significant
when the charge was murder, as those found GBMI
received life sentences 70.6 percent of the time, as
opposed to 49.2 percent of the time for non-mentally
ill defendants found guilty.
Another criticism of the GBMI verdict is that it
does not always guarantee mental health treatment
for the inmate. Some states make no provision for
mandatory treatment. For example, the statute in
Georgia specifically states that prisoners found
GBMI receive treatment only as financial resources
permit.15 In addition, in People v. Marshall16 the
Illinois Court of Appeals ruled that failure to assure
treatment for individuals found GBMI does not
make the statute unconstitutional.
Another criticism is that jurors may not understand the difference between the NGRI and GBMI
verdicts and may see a GBMI finding as a compromise verdict between guilty and NGRI.5 The GBMI
verdict gives jurors two verdicts (guilty and guilty but
mentally ill) by which to find a defendant guilty and
only one verdict (Not Guilty by Reason of Insanity)
by which to find the defendant not guilty when the
insanity defense is raised. In one study, prosecutors
agreed that the GBMI statute makes it easier to
prosecute a case when a defendant pleads insanity.17
The Debate About Juror Instructions
A common misperception held by the public is
that defendants found NGRI are released into the
community just as any other acquitted individual
would be.18 This has led some to argue that juries
Volume 33, Number 2, 2005
209
Juror Knowledge and Attitudes in Mental Illness Verdicts
should always be instructed that NGRI acquittees
will be committed for treatment and will not be released until they are judged not to be a threat to
society. In response, numerous courts have ruled on
whether a defendant has a right to have the jury
instructed as to the consequences of an NGRI verdict. A small number of states have approved the use
of instructions to a jury as to the outcome of an
NGRI verdict (California, Colorado, Florida, Louisiana, Massachusetts, Maryland, Pennsylvania, and
Utah).19 However, most courts have ruled that such
instructions would distract a jury from its sole function as the trier of fact and that consideration of the
disposition of these defendants would influence jury
decision-making. These courts have held that jurors
should only consider the statutory definitions of
these verdicts during their deliberation.
This was the case in South Carolina, as decided in
State v. Gary Allen Rimert.20 In this case, the defendant was found GBMI on a charge of murder and
appealed the decision, in part, on whether the jury
should have been instructed as to the dispositional
consequences of the GBMI and NGRI verdicts. The
Supreme Court of South Carolina upheld the trial
judge’s decision not to instruct the jury regarding the
dispositional consequences of these verdicts.
Study Hypotheses
This study was designed to test several hypotheses.
Mental illness verdicts are not easily understood and
are not part of common, everyday knowledge. Therefore, we hypothesized that most prospective jurors
could not correctly identify the definitions and dispositional outcomes of the NGRI and GBMI verdicts. We also hypothesized that jurors would consider the consequences (i.e., perceived dispositional
outcomes) of their verdicts and would let those beliefs influence their verdicts, even if instructed not to
do so by the trial judge. This study was designed to
measure jurors’ knowledge about two subjects: the
meaning of these verdicts in South Carolina and the
disposition of individuals found NGRI or GBMI. In
addition, we assessed juror attitudes as to what they
thought the disposition should be of defendants
found NGRI or GBMI.
Method
This study was approved by the Institutional Review Board of the University of South Carolina and
210
the Chief Administrative Judge for the Fifth Judicial
Circuit (Richland and Kershaw Counties). A written, multiple-choice instrument was given to volunteers from a qualified circuit court jury pool in Richland County, South Carolina, by the Clerk of Court
as they were discharged from jury duty. Residents of
Richland County who are called for jury duty serve
for an entire week. At the end of that week, our
questionnaires were handed out to discharged jurors,
whether they had participated in a jury trial or not. In
addition to the questionnaire, jurors were provided a
stamped, self-addressed envelope to return the survey. Richland County contains the state’s capital
(Columbia) and has a population of approximately
325,000 persons. The high school graduation rate is
85 percent and the median income is $40,000.
The survey questioned jurors about the following
demographic variables: sex, age, educational level,
occupation, and the juror’s relationship to the law
enforcement and legal communities (juror, friend,
family member, or none). Jurors were also asked if
they and/or anyone they knew “suffered from a severe mental illness such as Schizophrenia, Bipolar
Disorder (Manic Depression), or Major Depression,” with the same choice of response (juror, friend,
family member, or none).
The prospective jurors were asked to choose the
correct definition of the NGRI and GBMI verdicts
using the following choices: (1) the defendant has a
severe mental illness and cannot stand trial; (2) the
defendant was mentally ill at the time of the crime;
(3) the defendant had a mental illness that prevented
him or her from understanding that what he or she
did was wrong (correct response for NGRI); and (4)
the defendant had a mental illness that prevented
him or her from controlling his or her actions according to the law (correct response for GBMI). The
prospective jurors were asked to identify the dispositional outcome of the NGRI and GBMI verdicts and
what the outcomes should be, using the following
choices: (1) the defendant goes home; (2) the defendant goes to prison; (3) the defendant goes to a psychiatric hospital for treatment and is transferred to
prison when stable enough to complete the sentence
(correct response for GBMI); and (4) the defendant
goes to a psychiatric hospital and is then released to
go home when he or she is no longer a danger to himor herself or others (correct response for NGRI). The
prospective jurors were asked whether a defendant
The Journal of the American Academy of Psychiatry and the Law
Sloat and Frierson
Table 1
Demographic Variables
Variable
Sex
Educational level*
Relationship to law
enforcement†
Relationship to a lawyer†
Relationship to a judge†
Relationship to someone with
a major mental illness†
Table 2
Juror Knowledge
% Total Group
57.3 female
42.7 male
2.1 less than high school
12.5 high school graduate or GED
31.3 some college
27.1 college graduate
27.1 postgraduate education
2.0 self
13.5 friend
10.3 family member
77.1 none
4.1 self
21.8 friend
13.5 family member
62.5 none
0.0 self
7.3 friend
2.1 family member
90.6 none
6.2 self
8.3 friend
26.0 family member
67.7 none
The mean age was 46 years (range, 19 – 69).
* Values in this category do not total 100.0% due to rounding.
† Values in these categories do not total 100.0% as answers are not mutually
exclusive.
found guilty but mentally Ill could receive the death
penalty.
Finally, the prospective jurors were asked if they
thought that jurors should be instructed as to the
dispositional outcomes of the NGRI and GBMI verdicts before deliberation. They were also asked if this
knowledge would influence their verdict, even if the
judge instructed them not to consider dispositional
outcomes in reaching a verdict.
Results
Of the 200 surveys distributed to the qualified jury
pool, 101 (50.5%) were returned. Five responders
gave multiple answers to a single-answer question or
failed to respond to one or more questions. Therefore, 96 of the 101 questionnaires were included in
the analysis. The demographic variables are summarized in Table 1. The responding jurors were highly
educated, with 97.9 percent having at least a high
school education, and 54.2 percent having completed college and/or post-graduate education. Six
percent of jurors admitted to having a mental illness,
and 34 percent indicated that they knew someone
with a major mental illness. The occupational profile
of the responder group was diverse: 47 percent were
% Correct
% Incorrect
55.3
37.2
62.5
78.1
4.2
44.7
62.8
37.5
21.9
—
Definition of NGRI
Definition of GBMI
Dispositional outcome of NGRI
Dispositional outcome of GBMI
All of above
nonprofessionals, 35 percent were professionals, 16
percent were unemployed, and 2 percent did not
identify their occupations.
Juror knowledge is summarized in Table 2. Only
4.2 percent of all respondents identified the correct
meanings and outcomes of both the NGRI and
GBMI verdicts. When questioned as to the legal definition of the NGRI verdict, 55.3 percent of the
prospective jurors answered correctly, while 24.5
percent chose the definition of the GBMI verdict.
When questioned as to the legal definition of the
GBMI verdict, 37.2 percent answered correctly,
while 27.7 percent chose the definition of the NGRI
verdict.
Regarding the disposition of an NGRI verdict,
62.5 percent of responders identified the correct disposition; however, 27.1 percent chose the GBMI disposition (hospital, then prison), and 10.4 percent
believed the defendant would go home. Among jurors who chose the correct legal definition for NGRI,
71.2 percent also chose the correct disposition, while
15.4 percent chose the disposition for GBMI, and
13.5 percent answered that they believed the defendant would go home (Table 3).
When asked about the dispositional outcome of
the GBMI verdict, 78.1 percent answered correctly,
and 17.7 percent chose the disposition of those
found NGRI. Among jurors who chose the correct
legal definition of a GBMI verdict, 88.6 percent
Table 3 Responses Regarding Dispositional Outcome of NGRI
Among Those Who Knew Correct Legal Definition
Dispositional Dispositional
Outcome
Outcome
Is, %
Should Be, %
Psychiatric hospital, then home when
no longer dangerous (correct
answer)
Psychiatric hospital, then prison once
stabilized
Home
Prison
71.2
56.9
15.4
39.2
13.5
0.0
2.0
2.0
Columns do not total 100.0% due to rounding.
Volume 33, Number 2, 2005
211
Juror Knowledge and Attitudes in Mental Illness Verdicts
Table 4 Responses Regarding Dispositional Outcome of GBMI
Among Those Who Knew Correct Legal Definition
Dispositional Dispositional
Outcome
Outcome
Is, %
Should Be, %
Psychiatric hospital, then home when
no longer dangerous
Psychiatric hospital, then prison once
stabilized (correct answer)
Home
Prison
11.4
8.6
88.6
80.0
0.0
0.0
2.9
8.6
Columns do not total 100.0% due to rounding.
chose the correct outcome, while 11.4 percent chose
the outcome for an NGRI verdict. The majority of
respondents (68.4%) answered incorrectly, believing
that a defendant found GBMI could not receive a
death sentence (Table 4).
In assessing juror attitude about the disposition of
NGRI acquittees, 56.8 percent of jurors who identified the correct legal definition of NGRI believed
that defendants should go to a hospital and then
home when they were no longer dangerous, while
39.2 percent thought that defendants should go to
prison after being hospitalized. Jurors were more unified in their attitudes about the disposition of defendants found GBMI. Eighty percent of jurors who
correctly identified the legal definition of GBMI believed that the defendant should go to a hospital and
then to prison.
A large majority (84%) of respondents believed
that jurors should be informed of the outcome of
these verdicts before deliberation and 70.6 percent
reported that knowing the outcome would influence
their decisions, even if the judge instructed them not
to consider the outcome in arriving at a verdict.
There was little statistical correlation among the
demographic variables studied and knowledge of
mental illness verdicts and their dispositional outcomes. The only statistically significant finding was
that among those jurors with a post-high school education, those who did not complete a degree were
less likely to identify the outcome of a GBMI verdict
correctly ( p ⬍ .005).
Finally, among jurors who knew the correct legal
definition of NGRI, those with at least some college
education were more likely to think that insanity
acquittees should go to a hospital and then home
while jurors with a high school education or less believed that insanity acquittees should go to a hospital
and then prison ( p ⬍ .026). This difference did not
exist for the GBMI verdict.
212
Discussion
Among demographic variables, the educational
level of the responders was well above the 85 percent
high school graduation rate for this area. Less than
2.1 percent of the responders did not have a high
school education, and 55 percent had a college degree. This probably reflects sampling bias. Because
the survey involved a written instrument, those jurors who are illiterate may have been excluded. In
addition, because this study was presented as a
project of the University of South Carolina (located
in Richland County), those associated with that institution may have been more likely to complete the
study. Finally, because these surveys were distributed
to jurors during the summer months, the jury pool
may have been more educated, because professors,
teachers, and students may defer jury duty to the
summer months. Despite the high educational level
of this sample, only 4.2 percent of the prospective
jurors correctly identified both the legal definitions
and dispositional outcomes of the NGRI and GBMI
verdicts. This finding was not surprising to us, as
these are difficult concepts and are not part of common, everyday knowledge or general education.
Six percent of the responders reported having a
major mental illness (defined in the question as
Schizophrenia, Bipolar Disorder, or Major Depressive Disorder). Thus, the point prevalence of major
mental illness in this sample may be higher than in
the general population and may also represent sampling bias, as mentally ill individuals may have been
more interested in taking the survey.
The lack of significant correlation between demographic variables and knowledge base was consistent
across age, sex, relationship to law enforcement, relationship to someone in the legal community, and
relationship to someone with a major mental illness.
The finding that, compared with jurors with a college or post-graduate degree, those who attended college but did not complete a degree were less likely to
identify the outcome of the GBMI verdict correctly
( p ⬍ .05) is probably true of jurors with a high school
diploma or less. However, this finding was probably
obscured in groups in the lower educational levels by
the low response rate. Therefore, the difference did
not reach statistical significance. Although knowledge about mental illness verdicts is poor across all
educational levels, a higher educational level may
predict better knowledge about dispositional out-
The Journal of the American Academy of Psychiatry and the Law
Sloat and Frierson
comes of the NGRI and GBMI verdicts. Among jurors who correctly identified the meaning of NGRI,
juror attitude about the disposition of mentally ill
defendants appears to be influenced by educational
level. The less educated jurors were significantly
more punitive, believing that those found NGRI
should eventually go to prison after hospitalization
rather than home when no longer dangerous.
In the wake of the recent U.S. Supreme Court
decision barring the execution of the mentally retarded,21 the finding that a majority of jurors did not
believe that a defendant found GBMI could receive a
death sentence may become an appellate issue. If jurors believe that GBMI individuals may not be executed, they may be more likely to decide a defendant
is GBMI rather than NGRI in capital cases if they
base their decision on perceived dispositional outcome. Because both the NGRI and GBMI verdicts
require a unanimous jury finding, only one juror
with such a misperception could be crucial in deciding the ultimate outcome in a capital case. The existence of GBMI inmates on death row may become
the next battleground in the debate over capital
punishment.
In this study, we found that a large majority (84%)
of prospective jurors wanted to know the dispositional outcomes of mental illness verdicts and would
consider these outcomes when reaching a verdict,
even if instructed not to by the trial judge (70.6%).
These results suggest that jurors may decide verdicts
partly on what they believe will be the dispositional
outcome of the verdict. Because only 4.2 percent of
prospective jurors in this study correctly identified
the meaning and disposition of both the NGRI and
GBMI verdicts, it can be hypothesized that jurors
may be making decisions based on erroneous perceptions. An argument can be made that jurors should
be provided with accurate information regarding dispositional outcomes prior to jury deliberation.
References
1.
2.
3.
4.
5.
6.
7.
8.
9.
10.
11.
12.
13.
14.
15.
16.
17.
18.
19.
20.
21.
Durham v. U.S., 214 F.2d 862 (D.C. Cir. 1954)
M’Naghten’s Case, 8 Eng. Rep. 718 (H.L. 1843)
Model Penal Code § 4.01 (1985)
People v. McQuillan, 221 N.W.2d 569 (Mich. 1974)
Blunt LW, Stock HV: Guilty but mentally ill: an alternative verdict. Behav Sci Law 3:49 – 68, 1985
Mich. Comp. Laws Ann. § 768.36(1) (West 2003)
Miller RD: Criminal responsibility, in Principles and Practice of
Forensic Psychiatry (ed 2). Edited by Rosner R. London: Arnold
Publishers, 2003, pp 213–32
S.C. Code Ann. §17-24-10 (A) (Law Co-op. 2003)
S.C. Code Ann. §17-24-20 (A) (Law Co-op. 2003)
American Psychiatric Association: Position paper on the insanity
defense. Available at http://www.psych.org. Accessed May 20,
2004
American Psychological Association: Council Policy Manual: N.
Public Interest: Part 3. Available at http://www.apa.org. Accessed
May 20, 2004
American Bar Association: Mental health, mental retardation, and
criminal justice: general professional obligations. Available at
http://www.abanet.org. Accessed May 20, 2004
Petrella RC, Benedek EP, Bank SC, et al: Examining the application of the guilty but mentally ill verdict in Michigan. Hosp
Community Psychiatry 36:254 –9, 1985
Callahan LA, McGreevy MA, Cirincione C, et al: Measuring the
effects of the guilty but mentally ill (GBMI) verdict: Georgia’s
1982 GBMI reform. Law Hum Behav 16:447– 62, 1992
Ga. Code Ann. § 17-7-131 (2003)
People v. Marshall, 448 N.E.2d 969, 980 (Ill. Ct. App. 1983)
Klofas J, Yandrastis J: “Guilty but mentally ill” and the jury trial:
a case study. Crim Law Bull 24:423– 43, 1988
Schwartz BR: Should juries be informed of the consequences of
the insanity verdict? J Psychiatry Law 8:167– 89, 1980
Wheatman SR, Shaffer DR: On finding for defendants who plead
insanity: the crucial impact of dispositional instructions and opportunity to debate. Law Hum Behav 25:167– 83, 2001
State v. Gary Allen Rimert, 446 S.E.2d 400 (S.C. 1994)
Atkins v. Virginia, 536 U.S. 304 (2002)
Volume 33, Number 2, 2005
213